See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: NousResearch/CodeLlama-13b-hf
bf16: true
chat_template: llama3
datasets:
- data_files:
  - 4c22d7fb0719cc8f_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/4c22d7fb0719cc8f_train_data.json
  type:
    field_instruction: ' A painter needed to paint 12 rooms in a building. Each room
      takes 7 hours to paint. If he already painted 5 rooms, how much longer will
      he take to paint the rest? '
    field_output: x=(7.0*(12.0-5.0))
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 5
eval_table_size: null
flash_attention: false
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: false
group_by_length: false
hub_model_id: lesso11/0c44f59f-67de-4027-8a12-80d6757dc187
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: true
local_rank: null
logging_steps: 1
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 8
lora_target_linear: true
lr_scheduler: cosine
max_steps: 25
micro_batch_size: 2
mlflow_experiment_name: /tmp/4c22d7fb0719cc8f_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 1
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 10
sequence_len: 512
special_tokens:
  pad_token: </s>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 7d9074c3-15cf-4768-a2d9-16e4f969e782
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 7d9074c3-15cf-4768-a2d9-16e4f969e782
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

0c44f59f-67de-4027-8a12-80d6757dc187

This model is a fine-tuned version of NousResearch/CodeLlama-13b-hf on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1660

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0002
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_BNB with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• training_steps: 25

Training results

Training Loss	Epoch	Step	Validation Loss
9.4017	0.0017	1	2.1967
8.4538	0.0084	5	2.0898
4.0691	0.0169	10	0.5549
0.6929	0.0253	15	0.2638
0.1879	0.0337	20	0.1788
0.944	0.0422	25	0.1660

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1